A Facile Approach to Architecturally Defined Nanoparticles via

Jun 28, 2002 - A novel approach is presented for the controlled intramolecular collapse of linear polymer chains to give well-defined single-molecule ...
6 downloads 10 Views 177KB Size
Published on Web 06/28/2002

A Facile Approach to Architecturally Defined Nanoparticles via Intramolecular Chain Collapse Eva Harth,† Brooke Van Horn, Victor Y. Lee, David S. Germack, Chad P. Gonzales, Robert D. Miller, and Craig J. Hawker* Contribution from the Center for Polymeric Interfaces and Macromolecular Assemblies, IBM Almaden Research Center, 650 Harry Road, San Jose, California, 95120-6099 Received March 15, 2002

Abstract: A novel approach is presented for the controlled intramolecular collapse of linear polymer chains to give well-defined single-molecule nanoparticles whose structure is directly related to the original linear polymer. By employing a combination of living free radical polymerization and benzocyclobutene (BCB) chemistry, nanoparticles can be routinely prepared in multigram quantities with the size being accurately controlled by either the initial degree of polymerization of the linear chain or the level of incorporation of the BCB coupling groups. The latter also allows the cross-link density of the final nanoparticles to be manipulated. In analogy with dendritic macromolecules, a significant reduction of up to 75% in the hydrodynamic volume is observed on going from the starting random coil linear chains to the corresponding nanoparticles. The facile nature of the living free radical process also permits wide variation in monomer selection and functional group incorporation and allows novel macromolecular architectures to be prepared. Furthermore, the use of block copolymers functionalized with benzocyclobutene groups in only one of the blocks gives, after intramolecular collapse, a hybrid architecture in which a single linear polymer chain is attached to the globular nanoparticle.

Introduction

The synthesis and application of polymeric nanoparticles has attracted significant attention in recent years with further refinement of traditional methods as well as novel strategies for their preparation being developed.1-3 One of the driving forces for this interest has been the realization that functionalized nanoparticles can be considered as building blocks for a variety of nanotechnological applications, ranging from vectors for drug and DNA delivery systems4 to templating agents for nanoporous microelectronic materials.5 The strategies for preparing nanoparticles can be broadly classified into two main approaches, a top-down approach where emulsion polymerization techniques are further refined and optimized, leading to the development of microemulsion procedures resulting in particles from 20 to 50 nm.6 More recently, bottom-up techniques have been introduced which either rely on the synthesis of discrete * To whom correspondence should be addressed. E-mail: hawker@ almaden.ibm.com. † Present address: Xenoport Inc., 3410 Central Expressway, Santa Clara, CA 95051. (1) Bergbreiter, D. E. Angew. Chem, Int. Ed. 1999, 38, 2870. (2) (a) Monteiro, M. J.; Bussels, R.; Wilkinson, T. S. J. Polym. Sci., Part A: Polym. Chem. 2001, 39, 2813. (b) Park, M. K.; Xia, C.; Advincula, R. C. Langmuir 2001, 17, 7670. (3) (a) Thurmond, K. B., II; Kowalewski, T.; Wooley, K. L. J. Am. Chem. Soc. 1997, 119, 6656. (b) Demers, L. M.; Park, S. J.; Taton, T. A.; Li, Z.; Mirkin, C. A. Angew. Chem, Int. Ed. 2001, 40, 3071. (4) (a) Liu, M.; Fre´chet, J. M. J. Pharm. Sci. Technol. Today 1999, 2, 393. (b) Liu, J.; Zhang, Q.; Remsen, E. E.; Wooley, K. L. Biomacromolecules 2001, 2, 362. (c) Niemeyer, C. M. Angew. Chem., Int. Ed. 2001, 40, 4128. (5) (a) Hedrick, J. L.; Miller, R. D.; Hawker, C. J.; Carter, K. R.; Volksen, W.; Yoon, D. Y.; Trollsas, M. AdV. Mater. 1998, 10, 1049. (b) Hawker, C. J.; Hedrick, J. L.; Miller, R. D.; Volksen, W. MRS Bull. 2000, 25, 54. (c) Yang, S.; Mirau, P. A.; Pai, C. S.; Nalamasu, O.; Reichmanis, E.; Lin, E. K.; Lee, H. J.; Gidley, D. W.; Sun, J. N. Chem. Mater. 2001, 13, 2762. 10.1021/ja026208x CCC: $22.00 © 2002 American Chemical Society

spherical macromolecules such as dendrimers (1-10 nm)7 or the self-assembly of linear block copolymers into polymeric micelles followed by chemical cross-linking to give nanoparticles with typical dimensions ranging from 20 to 200 nm.8 As a consequence, the ability to routinely prepare nanoparticles in the 5-20 nm size range is limited. To address this issue, a new strategy involving the collapse and intramolecular coupling of single-polymer chains to give discrete nanoparticles has been proposed.9,10 While promising, this strategy has drawbacks, primarily the competing and statistically favored intermolecular cross-linking reaction which necessitated the use of ultra-dilute reaction conditions (ca. 10-510-6 M). This precludes the synthesis of these nanoparticles on a useful (>multigram) scale. In addition, even at these ultra(6) (a) Candau, F. In Microemulsions: Fundamental and Applied Aspects; Kumar, P., Mittal, K. L., Eds.; Marcel Dekker: New York, 1998. (b) Zhang, G.; Niu, A.; Peng, S.; Jiang, M.; Tu, Y.; Li, M.; Wu, C. Acc. Chem. Res. 2001, 34, 249. (c) Matyjaszewski, K.; Qiu, J.; Tsarevsky, N. V.; Charleux, B. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 4724. (d) Landfester, K. AdV. Mater. 2001, 13, 765. (e) Pilcher, S. P.; Ford, W. T. J. Polym. Sci., Part A: Polym. Chem. 2001, 39, 519. (7) (a) Bosman, A. W.; Jansen, H. M.; Meijer, E. W. Chem. ReV. 1999, 99, 1665. (b) Hecht, S.; Fre´chet, J. M. J. Angew. Chem., Int. Ed. 2001, 40, 74. (c) Percec, V.; Cho, W. D.; Ungar, G.; Yeardley, D. J. P. J. Am. Chem. Soc. 2001, 123, 1302. (d) Moore, J. S. Acc. Chem. Res. 1997, 30, 402. (e) Topp, A.; Bauer, B. J.; Klimash, J. W.; Spindler, R.; Tomalia, D. A.; Amis, E. J. Macromolecules 1999, 32, 7226. (8) (a) Huang, H.; Remsen, E. E.; Kowalewski, T.; Wooley, K. L. J. Am. Chem. Soc. 1999, 121, 3805. (b) Ma, Q.; Wooley, K. L. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 4805. (c) Liu, S.; Armes, S. P. J. Am. Chem. Soc. 2001, 123, 9910. (9) (a) Antonietti, M.; Bremser, W.; Pakula, T. Acta Polym. 1995, 46, 37. (b) Davankov, V. A.; Ilyin, M. M.; Tsyurupa, M. P.; Timofeeva, G. I.; Dubrovina, L. V. Macromolecules 1996, 29, 8398. (c) Martin, J. E.; Eichinger, B. E. Macromolecules 1983, 16, 1345. (d) Martin, J. E.; Eichinger, B. E. Macromolecules 1983, 16, 1350. (10) Mecerreyes, D.; Lee, V.; Hawker, C. J.; Hedrick, J. L.; Wursch, A.; Volksen, W.; Magbitang, T.; Huang, E.; Miller, R. D. AdV. Mater. 2001, 13, 204. J. AM. CHEM. SOC. 2002, 124, 8653-8660

9

8653

ARTICLES

Harth et al.

dilute conditions intermolecular cross-linking was still evident which results in poorly defined materials and in some cases gelation. To overcome this difficulty a continuous addition strategy for the successful synthesis of discrete nanoparticles by intramolecular cross-linking has now been developed. This permits the full potential of this strategy for the synthesis of well-defined and functionalized nanoparticles to be achieved. Furthermore the polymeric nanoparticles that are obtained from this procedure are freely soluble in common solvents and do not need surfactants, during either their synthesis or the subsequent stabilization of the resulting nanoparticle solutions. In many respects these new nanoparticle systems resemble dendrimers, although larger molecular sizes may be more easily prepared and they have the potential to be available in greater quantity. Experimental Section General Methods. Commercial reagents were obtained from Aldrich and used without further purification. Analytical TLC was performed on commercial Merck plates coated with silica gel GF254 (0.24 mm thick). Silica gel for flash chromatography was Merck Kieselgel 60 (230-400 mesh, ASTM). Nuclear magnetic resonance was performed on a Bruker AVANCE 400 FT-NMR spectrometer using deuterated solvents and the solvent peak as a reference. Gel permeation chromatography was performed in tetrahydrofuran (THF) on a Waters chromatograph equipped with four 5-µm Waters columns (300 mm × 7.7 mm) connected in series with increasing pore size (100, 1000, 100,000, 1,000,000 Å). A Waters 410 differential refractometer and a 996 photodiode array detector were employed. The polystyrene molecular weights were calculated relative to linear polystyrene standards, whereas the poly(n-butyl acrylate) molecular weights were calculated relative to poly(n-butyl acrylate) standards. 3-Carboxaldehydebicyclo[4,2,0]octa-1,3,5-triene or 4-Carboxaldehydebenzocyclobutene, 6. To a 500-mL flask was added 50 mL dry of THF, Mg turnings (2.88 g, 120 mmol), and 1,2-dibromoethane (4 drops). The reaction mixture was then heated under reflux for 15 min, 4-Bromobenzocyclobutene, 5,11 (20.0 g, 109 mmol) in 25 mL THF was added via a dropping funnel to form the Grignard reagent. After addition and rinsing the dropping funnel with 25 mL of dry THF, the reaction mixture was heated for an additional 45 min under reflux to give a green brown solution. The reaction mixture was then cooled to 0 °C, DMF (15 mL, 210 mmol) was added dropwise to the solution, and the reaction mixture was heated under reflux for 15 min. The reaction mixture was poured onto 150 g of ice, acidified to pH ) 4, and neutralized with saturated NaHCO3 solution. The crude product was extracted with ethyl acetate, the organic phase was filtered over Celite, and evaporation of the solvent gave the crude product. The product was purified by column chromatography using 10% diethyl ether/hexane as eluting solvents and was finally purified by Kugelrohr distillation (145 °C, 0.5 mm) to give the aldehyde, 6, (11.7 g, 81.2%) as a colorless liquid; IR 3000-2800, 1690, 1598, 1216, 1067 and 827 cm-1; 1H NMR (400 MHz, CDCl3) δ 9.9 (s, 1H, CHO), 7.65 (dd, 1H, J ) 7.4 Hz, J′ ) 1.2 Hz, ArH), 7.50 (s, 1H, ArH), 7.14 (dd, 1H, J ) 7.4 Hz, J′ ) 1.2 Hz, ArH), 3.15 (s, 4H, CH2); 13C NMR (100 MHz, CDCl3) δ 192.28, 153.69, 146.57, 135.4, 130.26, 122.89, 122.81, 29.97, and 29.23. Anal. Calcd for C9H8O; C, 81.8; H 6.10. Found: C, 81.7; H, 5.94. 3-Ethenylbicyclo[4,2,0]octa-1,3,5-triene or 4-Vinylbenzocyclobutene, 4. To a 500-mL round-bottom neck flask was added (Ph)3PCH3Br (24.3 g, 68.1 mmol), 110 mL of dry THF, and the solution was cooled to -78 °C. n-BuLi (2.5 M in hexane, 26.4 mL, 66 mmol) was added dropwise, and the reaction mixture was allowed to warm to room temperature The yellow-orange solution was cooled to -78 °C, and (11) Lloyd, J. B. F.; Ongley, P. A. Tetrahedron 1965, 21, 245. 8654 J. AM. CHEM. SOC.

9

VOL. 124, NO. 29, 2002

the aldehyde, 6 (7.16 g, 54.2 mmol), diluted in 34 mL of dry THF, was added slowly. The mixture warmed to room temperature, and stirring continued for 2 h. The reaction was treated sequentially with saturated NH4Cl and saturated NaHCO3 solution, and the crude product was filtered over Celite, washed with diethyl ether/hexane (1:1), and evaporated to dryness (no heat) to give the crude product. Further purification by column chromatography using 5% diethyl ether/hexane as an eluting solvent followed by Kugelrohr distillation (75 °C, 1.0 mm) gave the pure styrene derivative, 4, as a colorless liquid (5.50 g, 78%); IR 2925, 1627, 1473, 989, 901, and 829 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.26 (d, 1H, J ) 7.4 Hz, ArH), 7.20 (s, 1H, ArH), 7.04 (d, 1H, J ) 7.4 Hz, ArH), 6.74 (dd, 1H, J ) 17.5 Hz, J′ ) 10.8 Hz, CH), 5.70 (d, 1H, J ) 17.5 Hz, CH2), 5.20 (d, 1H, J ) 10.8 Hz, CH2), 3.19 (s, 4H, CH2); 13C NMR (100 MHz, CDCl3) δ 146.09, 145.75, 137.94, 136.69, 125.71, 122.58, 119.90, 112.38, 29.52, and 29.35. Anal. Calcd for C10H10; C, 92.2; H 7.80. Found: C, 92.0; H, 8.03. Random Copolymer of 4 and Styrene, 8. The alkoxyamine initiator, 7 (32.5 mg, 0.1 mmol),12 dissolved in styrene (10.4 g, 100 mol) and 4-vinylbenzocyclobutene, 4 (3.25 g, 25.0 mmol), was added to a glass ampule with a stir bar. After three freeze and thaw cycles the ampule was sealed under argon and heated for 6 h at 120 °C. The resulting polymer was dissolved in dichloromethane and purified by precipitation into a 1:1 mixture of 2-propanol/acetone followed by reprecipitation into methanol to give 8 as a colorless powder (12.1 g, 88%), Mw ) 111 000; PDI ) 1.11; IR 3100-2850, 1601, 1492, 1452, 909, and 699 cm-1; 1H NMR (400 MHz, CDCl3) δ 7.24-6.57 (m, ArH), 3.05 (br s, CH2), 1.83-1.26 (m, CH2, CH); 13C NMR (100 MHz, CDCl3) δ 145.0-146.4, 1127.9, 125.5, 121.8, 42.0-44.0, 40.4, and 29.2. Random Copolymer of 4 and n-Butylacylate, 10. The alkoxyamine initiator 7 (32.5 mg, 0.1 mmol) was dissolved in n-butyl acrylate (10.2 g, 72.0 mmol) and 4 (1.04 g, 8.0 mmol) and placed in a glass ampule with a stir bar. After three freeze and thaw cycles the ampule was sealed under argon and heated for 15 h at 125 °C. The resulting polymer was dissolved in dichloromethane and precipitated in MeOH/H2O (3:1) to give 10 as a colorless gum (10.2 g, 91%), Mw ) 77 500; PDI ) 1.12; 1 H NMR (400 MHz, CDCl3) δ 6.83-6.63 (m, ArH), 4.10-3.83 (m, CH2,CH), 3.05 (bs, CH2), 2.22-1.01 (m, CH2, CH3). Methyl(2,2,5-Trimethyl-3-(benzylethoxy)-4-phenyl-3-azahexane)poly(ethylene Glycol), 12. NaH (0.23 g, 6.3 mmol) was slowly added to a mixture of monomethylpoly(ethylene glycol), 14 (7.85 g, 1.57 mmol), and 18-crown-6 (10 mg) dissolved in 10 mL of THF under a constant argon flow. After 15 min, the chloromethyl-substituted alkoxyamine, 13 (1.16 g, 3.14 mmol)12 was added to the reaction mixture, which was subsequently heated at reflux for 16 h. After the addition of a few drops of water to neutralize the excess NaH, the reaction mixture was concentrated, dissolved in dichloromethane, filtered, and evaporated to dryness. The crude product was obtained after flash chromatography eluting with dichloromethane gradually increasing to 10% methanol/dichloromethane to give the PEG-macroinitiator, 12, as a colorless solid (8.03 g, 89%); IR (KBr) 3439 cm-1 (NH), 1693 cm-1 (amide).1H NMR (400 MHz, CDCl3) δ 7.4-7.1 (m, ArH), 5.10 (d, CH), 4.92 (d, CH2OAr), 3.65 (s, OCH2), 3.41 (d, CH), 3.28 (d, CH), 2.43 (m, CH), 1.65 (d, CH3), 1.52 (d, CH3), 1.40 (m, CH), 1.33 (d, CH3), 1.05 (s, t-Bu), 0.89 (d, CH3), 0.80 (s, t-Bu), 0.61 (d, CH3), and 0.22 (d, CH3). Poly(ethylene glycol)-b-(styrene-co-benzocyclobutene), 15. The poly(ethylene glycol) terminated alkoxyamine, 12 (500 mg, 0.1 mmol) (Mn ) 5 000, PDI ) 1.06) was dissolved in styrene (10.4 g, 100 mol) and 4-vinylbenzocyclobutene, 4 (3.25 g, 25.0 mmol) in a glass ampule with a stir bar. After three freeze and thaw cycles the ampule was sealed under argon and heated for 6 h at 125 °C. The resulting polymer was (12) (a) Benoit, D.; Chaplinski, V.; Braslau, R.; Hawker, C. J. J. Am. Chem. Soc. 1999, 121, 3904. (b) Dao, J.; Benoit, D.; Hawker, C. J. J. Polym. Sci., Part A: Polym. Chem. 1998, 36, 2161. (c) Hawker, C. J.. J. Am Chem. Soc. 1994, 116, 11185.

Nanoparticles via Intramolecular Chain Collapse dissolved in dichloromethane and purified by precipitation into a 1:1 mixture of isopropoanol/acetone followed by reprecipitation into methanol to give 15 as a colorless powder (10.7 g, 76.1%), Mw ) 89 500; PDI ) 1.12; 1H NMR (400 MHz, CDCl3) δ 7.24-6.57 (m, ArH), 3.65 (s, OCH2), 3.05 (br s, CH2), 1.83-1.26 (m, CH2, CH). General Procedure for Nanoparticle Formation, 9. In a 500-mL three-necked flask equipped with a internal thermometer, condenser, and septum, 120 mL of benzyl ether was heated at 250 °C under argon. A solution of the benzocyclobutene (BCB)-functionalized linear polymer, 8 (4.00 g, Mn ) 108 000; PDI ) 1.15, 7.5 mol % BCB), dissolved in benzyl ether (40 mL) was added dropwise via a peristaltic pump at ca. 12.8 mL/h with vigorously stirring under argon. After addition the reaction mixture was heated for an additional 1 h, the solvent was distilled under reduced pressure, and the remaining crude product was dissolved in dichloromethane and precipitated into methanol. This gave the nanoparticles, 9, as a colorless solid (3.76 g, 94% yield), 1H NMR (400 MHz, CDCl3). The significant change is the disappearance of the aliphatic benzocylobutene protons at 3.05 on formation of the cross-linked nanoparticles; all other aspects of the spectrum are similar.

ARTICLES Scheme 1. Coupling Reaction of Benzocyclobutene Derivatives, 1.

Scheme 2. Synthesis of Benzocyclobutene Monomer, 4.

Results and Discussion

In developing a new strategy for nanoparticle formation, we were drawn to the field of living free radical polymerizations where the high degree of control is a result of an equilibrium between dormant and reactive propagating radicals. As a result, the reactive radical chain ends are present in extremely low concentrations.13 This general concept, that is, only the reactive species need to be at ultra-dilute concentrations, was then applied to the formation of nanoparticles by intramolecular coupling. In this case, the linear polymer, which contains numerous latent coupling groups along the backbone, is added slowly to a heated solvent in which the coupling groups are either thermally or chemically activated. As a consequence, the traditional conditions of ultrahigh dilution need only be met for the reactive intermediates and not for the polymers themselves. Following this coupling event, the nanoparticles should be unreactive, which allows their concentration to increase to very high levels, (0.1-1.0 M) without intermolecular cross-linking reactions leading to gelation or coupling of individual nanoparticles. The ability to work at 0.1-1.0 M concentrations can be compared to the impractical concentration levels (ca. 10-6 M) required for traditional ultrahigh dilution techniques. To satisfy these demands, the nature of the cross-linking group is critical; it must be selectively activated and react rapidly, leading to efficient intramolecular bond formation. It is also important that this reaction is irreversible and leads to a coupled structure that is subsequently unreactive under the reaction conditions. To fulfill these goals, our attention was directed to the benzocyclobutene group (BCB) which has found wide use as a latent Diels-Alder reagent in organic synthesis14 (13) (a) Matyjaszewski, K.; Qiu, J.; Tsarevsky, N. V.; Charleux, B. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 2440. (b) Stalmach, U.; de Boer, B.; Post, A. D.; van Hutten, P. F.; Hadziioannou, G. Angew. Chem., Int. Ed. 2001, 40, 428. (c) Percec, V.; Barboiu, B.; Bera, T. K.; van der Sluis, M.; Grubbs, R. B.; Fre´chet, J. M. J. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 4776. (d) Bon, S. A. F.; Steward, A. G.; Haddleton, D. M. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 2678. (e) Tokuchi, K.; Ando, T.; Kamigaito, M.; Sawamoto, M. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 4735. (f) Benoit, D.; Hawker, C. J.; Huang, E. E.; Lin, Z.; Russell, T. P. Macromolecules 2000, 33, 1505. (g) Becker, M. L.; Remsen, E. E.; Wooley, K. L. J. Polym. Sci., Part A: Polym. Chem. 2001, 39, 4152. (h) Moschogianni, P.; Pispas, S.; Hadjichristidis, N. J. Polym. Sci., Part A: Polym. Chem. 2001, 39, 650. (i) Rodlert, M.; Harth, E.; Rees, I.; Hawker, C. J. J. Polym. Sci. Polym. Chem. 2000, 38, 4749. (14) Segura, J. L.; Martin, N. Chem. ReV. 1999, 99, 3199.

and in the formulation of thermosetting materials.15 Upon heating, the benzocyclobutene group, 1, undergoes ring opening to give the extremely reactive o-quinoid structure, 2, which primarily reacts via an irreversible dimerization reaction to form the dibenzocyclooctadiene derivative, 3, as well as a mixture of unidentified oligomeric materials (Scheme 1). The direct result of this chemistry is the selective formation of cross-links from the coupling of two or more benzocyclobutene units.16 Having identified the benzocyclobutene functionality as the critical coupling group for the preparation of nanoparticles, the desired monomer, 4-vinylbenzocyclobutene, 4, was prepared from 4-bromobenzocyclobutene, 5, by initial Grignard formation followed by reaction with N,N-dimethylformamide to give the aldehyde, 6. Wittig coupling of 6 with methyltriphenylphosphonium bromide afforded the desired styrene derivative, 4, in high yield (Scheme 2). As anticipated, the incorporation of the cyclobutene group into the monomer, 4, did not decrease its stability when compared to styrene, and 4 proved to be stable to a wide variety of reaction conditions. The key to the success of this intramolecular chain collapse strategy is the elimination of intermolecular cross-linking between BCB groups on different chains. Since it was envisaged that this critical balance between intramolecular coupling and intermolecular cross-linking would be influenced by the number and placement of BCB units along the polymeric backbone, accurate control of the starting linear polymers structure was also considered important. To achieve this and to permit a wide variety of linear polymers to be conveniently prepared, the polymerization of the desired monomer, 4-vinylbenzocyclobutene, 4, was examined under living free radical conditions.17 Copolymerization of 4 with vinyl monomers such as styrene, methyl methacrylate, or n-butyl acrylate in the presence of the R-hydrido alkoxyamine, 7, proved to be a controlled (15) (a) Deeter, G. A.; Venkataraman, D.; Kampf, J. W.; Moore, J. S. Macromolecules 1994, 27, 2647. (b) Kraus, A.; Gugel, A.; Belik, P.; Walter, M.; Mu¨llen, K. Tetrahedron 1995, 51, 9927. (c) Kirchhoff, R. A. U.S. Patent 4,540,763, 1985. (16) Jensen, F. R.; Coleman, W. E.; Berlin, A. J. Tetrahedron Lett. 1962, 15. J. AM. CHEM. SOC.

9

VOL. 124, NO. 29, 2002 8655

ARTICLES

Harth et al.

Scheme 3. Synthesis of Benzocyclobutene Functionalized Linear Polystyrene, 8.

Figure 1. Variation in molecular weight of final macromolecule, Mw, with total concentration of BCB groups in solution; (0) ultra-high dilution strategy; (0) continuous addition strategy; (I) represents that an insoluble gel was produced (starting linear polymer; Mw ) 95 000; PDI ) 1.11).

procedure, leading to random incorporation of the reactive BCB units and low polydispersities for the resulting copolymers, 8 (Scheme 3). At molecular weights less than 120 000 amu the polydispersities for these random copolymers were 1.08-1.16, which increased to 1.19-1.26 for molecular weights above 200 000 amu. This increased polydispersity is due to the significantly reduced concentration of initiating groups at these high molecular weights and has been observed previously for both ATRP- and nitroxide-mediated procedures.17 The BCB-functionalized polystyrene derivatives, 8, could be readily characterized by standard techniques and incorporation of the BCB units monitored by 1H NMR which showed characteristic aliphatic resonances for the cyclobutene ring at 3.10 ppm. Intramolecular collapse was initially examined under traditional ultrahigh dilution techniques. For this, a solution of a 80:20 styrene/BCB random copolymer, 8 (Mw ) 95 000; PDI ) 1.11) in dibenzyl ether was heated at a variety of concentrations18 under N2 for 30 min at 250 °C. As can be seen in Figure 1, at very low concentrations of BCB groups, ca. 5.0 × 10-5 M (see inset, Figure 1), intermolecular cross-linking becomes (17) (a) Hawker, C. J.; Bosman, A. W.; Harth, E. Chem. ReV. 2001, 101, 36613688. (b) Kamigaito, M.; Ando, T.; Sawamoto, M. Chem. ReV. 2001, 101, 3689-3746. (c) Matyjaszewski, K.; Xia, J. Chem. ReV. 2001, 101, 29212990. (18) The concentration of BCB groups is an overall concentration in solution and is independent of the molecular weight of the linear polymer. For example, a 80/20 styrene/BCB copolymer has a repeat unit consisting of 4 styrene monomers (4 × 104) and one BCB monomer (1 × 130), which equals a repeat unit molecular weight of 546. Therefore 4.00 g of this 80/ 20 copolymer contains, 4.00/546 ) 7.32 × 10-3 mol of repeat units or 7.32 millimole equivalents of BCB. Dissolution in 50 mL of dibenzyl ether then leads to an overall concentration of 0.146 M. 8656 J. AM. CHEM. SOC.

9

VOL. 124, NO. 29, 2002

apparent as evidenced by the increase in molecular weight of the product from its base value of 65 000 for a discrete nanoparticle due to chain-chain coupling. This ability to readily identify chain-chain coupling has been previously observed for dendrimer chemistry where even minor amounts of intermolecular coupling can be easily detected by gel permeation chromatography (GPC). This ability is due to the molecular weight doubling on chain-chain coupling, and the combination of this feature with the low polydispersity of the initial chains results in a lower limit of ca. 1-2% of intermolecular cross-linking being readily detected as a higher molecular shoulder under standard GPC conditions. At higher concentrations18 of ca. 9.0 × 10-3 M crosslinking to a swollen gel occurs very rapidly due to the large number of BCB functional groups along the polymeric backbone. While these concentrations are comparable to results obtained with other traditional ultra-high dilution techniques,9,10 the results are in stark contrast to the continuous addition strategy. In this approach, a concentrated solution ([BCB] ) 0.2 M) of the same starting linear polymer, 8, is continuously added via a peristaltic pump to a high-boiling solvent, such as dibenzyl ether, heated at 250 °C19 to give a final BCB concentration of 0.05 M. After addition, the solvent is removed, and the nanoparticles, 9, are isolated using normal precipitation techniques (Scheme 4). No gelation or intermolecular crosslinking is observed under these conditions, and only after increasing the final concentration of BCB groups to 0.12 M were minor amounts of nanoparticle coupling observed. The ability to successfully conduct these chain-collapse reactions at final BCB concentrations of 0.01-0.1 M represents an increase of 3-4 orders of magnitude when compared to the traditional ultrahigh dilution strategy (Figure 1). This permits multigram samples to be prepared on a routine basis with standard laboratory equipment, a dramatic improvement compared to previous approaches. A significant feature of the above concentration studies is the reduction in hydrodynamic volume of the random coil linear polymer on intramolecular collapse to give the final nanoparticle. In the above example, the original linear polymer has a molecular weight, Mw of 95 000 amu; however, upon reaction the macromolecule decreases in size to give a nanoparticle with an apparent or polystyrene equivalent Mw of 65 000 amu. Dynamic light scattering was also employed to follow this decrease in size, and a reduction in the hydrodynamic radius, Rh from 8.7 to 6.6 nm, was observed upon intramolecular collapse. Since no byproducts are produced during this reaction (19) A reaction temperature of 250 °C corresponds to the exotermic maximum of the DSC curve relating to ring opening of the cyclobutene ring, the maximum was chosen to ensure efficient and rapid formation of the reactive o-quinoid intermediate.

Nanoparticles via Intramolecular Chain Collapse

ARTICLES

Scheme 4. Schematic Representation of the Intramolecular Collapse of the Linear Polymer, 8, To Give the Nanoparticle, 9.

Figure 2. Comparison of 1H NMR spectrum for (a) the starting linear polymer, 8, 80/20 Sty/BCB, Mw ) 95 000, PDI ) 1.12; and the resulting nanoparticle, 9, Mw ) 65 000, PDI ) 1.10.

and no molecular weight lost, this decrease can only be due to a change in the architecture of the macromolecule from a random coil to a nanoparticle. Again, this is consistent with dendrimer chemistry where the compact, three-dimensional dendritic structure leads to an apparent molecular weight which is significantly smaller that the actual molecular weight.20 Further confirmation of the structural change was obtained from NMR studies; of particular note is the observed absence of unreacted BCB units in the final nanoparticles, 9. As shown in Figure 2, comparison of the 1H NMR spectra for the starting linear polymer, 8, and the nanoparticle, 9, shows the prominent resonance for the aliphatic protons of the cyclobutene group at 3.10 ppm in the former, which completely disappear after collapse, and a broad resonance at 2.0-3.0 ppm is observed. This is consistent with ring opening of the benzocyclobutene group and coupling to give cyclooctane derivatives and higher aliphatic coupled oligomers. A direct consequence of this ring opening is that the BCB groups undergo reaction to give dimers and oligomers that do not undergo any further coupling chemistry. This fulfils one of the requirements discussed above for a successful intramolecular chain collapse reaction and permits the substantial build-up of product in the final reaction mixture. The lack of reactivity can also be demonstrated by repeated thermal cycling of the isolated nanoparticles, which results in no observable change in physical properties such as molecular weight, NMR spectra, and so forth. Formation of the nanoparticles also leads to an increase in the glass transition temperature of the nanoparticles when compared to the starting linear polymers. The BCB functionalized linear polystyrenes, 8, show Tg’s similar to that observed for polystyrene, ca. 100(20) (a) Fre´chet, J. M. J. Science 1994, 263, 1710. (b) Fischer, M.; Vo¨gtle, F. Angew. Chem., Int. Ed. 1999, 38, 885. (c) Hawker, C. J.; Fre´chet, J. M. J.. J. Am. Chem. Soc. 1990; 112, 7638. (d) Tomalia, D. A.; Naylor, A. M.; Goddard, W. A., III. Angew. Chem. Int. Ed. Engl. 1990, 29, 138.

Figure 3. Overlay of GPC traces for (a) the starting linear polymer, 8, Mw ) 105 000, PDI ) 1.12; and nanoparticles, 9, with (a) 0 mol % BCB incorporation, (b) 5 mol % BCB incorporation, (c) 10 mol % BCB incorporation, (d) 20 mol % BCB incorporation, and (e) 25 mol % BCB incorporation.

105 °C, while the glass transition temperature for the nanoparticles, 9, increase by ca. 20 °C to 120-130 °C at 20% BCB incorporation with an associated broadening of the transition. All of the above data is consistent with the intramolecular collapse of a random coil linear polymer to give a single, higherdensity nanoparticle. This unique feature of being able to tailor the nanoparticle via the starting linear polymer was then examined in detail using three different series of polystyrene derivatives, ca. 44 000; 110 000; and 230 000 amu containing varying levels of BCB incorporation from 1.25 to 30% (Table 1). Under the continuous addition technique described above, conversion of the linear polymers to nanoparticles was a facile process at all molecular weights and percent BCB incorporations studied. At concentrations of up to 0.01-0.1 M, no indication of intermolecular crosslinking was observed, and in each case the GPC trace shifted to lower hydrodynamic volumes. As can be seen in Figure 3, for the same molecular weight of the starting linear polymer, 8, a systematic decrease in the hydrodynamic volume of the nanoparticles is observed on increasing the percent of benzoJ. AM. CHEM. SOC.

9

VOL. 124, NO. 29, 2002 8657

ARTICLES

Harth et al.

Table 1. Comparison of the Polystyrene Equivalent Molecular Weights and PDI for the Starting Linear Polymers, 8, and the Final Nanoparticles, 9 linear

nanoparticle

composition

Mw

PDI

Mw

PDI

%BCB

Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB Sty/BCB

44 000 44 500 45 500 43 500 44 000 110 500 113 000 109 000 108 000 112 000 110 000 111 000 112 000 231 000 235 000 228 000 230 000 233 000 231 000 235 000 230 000 229 000 234 000

1.09 1.08 1.09 1.10 1.07 1.15 1.12 1.14 1.15 1.10 1.16 1.11 1.12 1.23 1.25 1.22 1.19 1.26 1.24 1.23 1.21 1.24 1.23

41 700 38 200 29 100 20 700 18 500 103 000 94 500 79 300 59 800 56 000 44 200 42 800 40 500 189 800 174 000 109 000 98 000 91 500 81 000 80 300 66 000 62 000 63 500

1.09 1.12 1.11 1.12 1.12 1.17 1.19 1.18 1.18 1.19 1.16 1.15 1.09 1.26 1.22 1.26 1.18 1.16 1.17 1.19 1.17 1.25 1.16

2.50 5.00 10.00 15.00 20.00 1.25 2.50 5.00 7.50 10.00 15.00 20.00 25.00 1.25 2.50 5.00 7.50 10.00 12.50 15.00 20.00 25.00 30.00

Figure 4. Variation in the percent reduction in molecular weight21 for the nanoparticles, 9, with the mol % of BCB units in the starting linear polymer, 8, for 44 K (b), 110 K (9), and 230 K (b) series. Table 2. Comparison of Molecular Weight (Linear Standard Equivalent) and PDI for the Starting Functionalized Linear Polymers, 10, and the Final Nanoparticles, 11 linear

cyclobutene groups, which is consistent with an increase in the level of intramolecular coupling and a more globular, threedimensional structure. The other pertinent feature of Figure 3b-e is the symmetrical nature/low PDI of the GPC traces for the nanoparticles and the associated lack of higher-molecular weight shoulders. This demonstrates that even at high BCB loadings, ca. 25 mol %, no detectable amount of intermolecular cross-linking is occurring. Analysis of the trends within each series showed that the percent reduction in hydrodynamic volume increases with both increasing molar percentage of BCB and the molecular weight of the starting linear polymer. In each case, the actual molecular weight of the cross-linked macromolecules is significantly greater than the apparent molecular weight. For example, a 70/ 30 styrene/BCB random copolymer with an initial molecular weight, Mw ) 234 000 (PDI ) 1.23) gives a nanoparticle with a polystyrene equivalent molecular weight, Mw ) 63 500 (PDI ) 1.16) which represents a reduction in hydrodynamic volume of 73%. The actual molecular weight of the final nanoparticle was also determined by light scattering and found to be 230 000 which is within experimental error of that of the starting linear polymer which demonstrates that the actual molecular weights of the starting linear polymer and the final nanoparticles are approximately the same. This collapse and associated change in hydrodynamic volume is therefore due to the formation of up to 310 intramolecular links per nanoparticle, assuming that each activated BCB group reacts with one other activated BCB group.22 Interestingly, all plots are of a similar shape and seem to reach a plateau of between 65 and 75% reduction in apparent molecular weight (Figure 4). It should also be noted that in the control experiments, heating polystyrene with 0% BCB incor(21) Percent reduction in molecular weight is calculated from the difference between the actual molecular weight and the apparent molecular weight as determined by GPC; (Mw actual - Mw apparent)/Mw actual × 100%, (22) This is based on the assumption of one BCB unit reacting with a second BCB unit forming an intramolecular dimer. While it is an efficient process, there is the possibility of side reactions which will affect the absolute numbers of cross-links; therefore, this number may represent an upper limit. 8658 J. AM. CHEM. SOC.

9

VOL. 124, NO. 29, 2002

nanoparticle

composition

Mw

PDI

Mw

PDI

%BCB

MMA/BCB MMA/BCB MMA/BCB n-BuA/BCB n-BuA/BCB n-BuA/BCB n-BuA/BCB Sty/Cl-Sty/BCBa Sty/Cl-Sty/BCBa Sty/Cl-Sty/BCBa PEG-Sty/BCBb PEG-Sty/BCBb PEG-Sty/BCBb

52 500 54 500 56 000 74 500 77 500 75 000 73 000 101 000 92 000 85 000 92 000 95 000 89 500

1.17 1.12 1.13 1.10 1.12 1.09 1.09 1.18 1.14 1.14 1.13 1.11 1.12

36 500 28 000 26 900 58 100 45 700 33 500 27 800 73 500 48 500 34 000 70 500 52 000 36 500

1.14 1.11 1.13 1.12 1.14 1.09 1.10 1.20 1.13 1.17 1.10 1.09 1.14

10.00 15.00 20.00 5.00 10.00 15.00 20.00 5.00 10.00 20.00 5.0 10.0 20.0

a 10 mol % incorporation of p-chloromethylstyrene ) 5000; PDI ) 1.06.

b

PEG block, Mn

Figure 5. GPC traces for (a) the starting poly(ethylene glycol)-b-poly(styrene-co-benzocyclobutene), 15, (Mw ) 95 000, PDI ) 1.11) and (b) the final hybrid nanoparticle-linear block copolymer, 16, (Mw ) 52 000, PDI ) 1.09).

poration resulted in no detectable change in the chromatographic or spectral properties of the polymers. Examination of the data in Table 1 also demonstrates the inherent versatility of this approach in controlling the size of the final nanoparticle. Not only can the size and cross-link density of the nanoparticle be controlled by the level of BCB incorporation, but the molecular weight of the starting linear

Nanoparticles via Intramolecular Chain Collapse

ARTICLES

Scheme 5. Schematic Representation of the Intramolecular Collapse of a Random n-Butyl Acrylate-Based Linear Polymer, 10, To Give the Nanoparticle, 11.

Scheme 6. Formation and Intramolecular Collapse of the PEG-b-PSt/BCB Block Copolymer, 15, To Give a Hybrid Linear-Nanoparticle Copolymer, 16.

polymer also plays a key role in determining the hydrodynamic volume of the final nanoparticle. For example, a polystyrene derivative with a 10% incorporation of BCB and a molecular weight, Mw, of 112 000 gives a nanoparticle with a Rh of 6.2 nm. Increasing the molecular weight, Mw, of the starting linear polymer to 233 000 while still retaining the 10% incorporation of BCB gives a larger nanoparticle with a Rh of 9.5 nm. In turn, a polystyrene derivative with an analogously higher molecular weight of 229 000 but with a 25% incorporation of BCB gives a nanoparticle with a Rh of 6.4 nm, very similar to the first example with a lower molecular weight (117 000) and level of BCB incorporation (10%). A consequence of this is that the size and physical characteristics of the final nanoparticle can be directly dictated by the structure and functionality of the starting linear polymer. The versatile nature of this intramolecular chain collapse approach to nanoparticles coupled with the ability to prepare a wide variety of linear polymers by living free radical techniques also opens up the possibility of preparing well-defined nanoparticles incorporating functional groups, nonstryrenic monomers, or different macromolecular architectures. As shown in

Table 2, starting linear polymers based on methyl methacrylate (MMA) or n-butyl acrylate (n-BuA) can be employed as the backbone polymer with no change in the efficiency of the intramolecular collapse process. For example, copolymerization of an 85:15 mixture of n-butyl acrylate and the vinyl BCB derivative, 4, in the presence of the alkoxyamine initiator, 7, proceeds smoothly to give the well-defined random copolymer, 10, with a molecular weight, Mw of 75 000 and a polydispersity of 1.09. Addition of a concentrated solution of 10 (0.1 M) to dibenzyl ether, heated at 250 °C gives a poly(n-butyl acrylate) nanoparticle, 11, with an apparent molecular weight, Mw of 33 500 and a polydispersity of 1.09. (Scheme 5). The relative selectivity of the thermal procedure used to activate the BCB group also allows other functional groups such as chloromethyl substituents to be introduced into the linear polymer, thereby leading to functionalized nanoparticles. It should however be realized that this intramolecular collapse procedure is not limited to simple linear random copolymers. Additional structural features can be built into the starting materials, which are then translated into the nanoparticle structure. For example, block copolymers can potentially be used in such an approach, and if the reactive BCB groups are contained in only one of the blocks, novel macromolecular architectures can be prepared in which a controlled number of linear chains, one or two for AB and ABA block copolymers respectively, are attached to the nanoparticle. To test this hypothesis, functionalized poly(styrene)-b-poly(ethylene glycol) J. AM. CHEM. SOC.

9

VOL. 124, NO. 29, 2002 8659

ARTICLES

Harth et al.

AB block copolymers were prepared by living free radical procedures. The alkoxyamine substituted poly(ethylene glycol) macroinitiator, 12, was obtained by reaction of the sodium salt of monomethylpoly(ethylene glycol), 13 (Mn ) 5000, PDI ) 1.06), with the chloromethyl substituted alkoxyamine, 14. The macroinitiator, 12, was then used to initiate the polymerization of a mixture of styrene and 4 at 120 °C to give the desired AB block copolymer, 15, which contains the cross-linking BCB units in the second block only (Scheme 6). Reaction of 15 under continuous addition conditions then results in the selective intramolecular collapse of the second block to give a novel hybrid linear-nanoparticle architecture, 16, in which a single water soluble PEG linear chain is attached to a three-dimensional cross-linked polystyrene nanoparticle, similar in structure to that of hybrid dendritic-linear block copolymers.23 As can be seen in Figure 5, the effect of intramolecular collapse is clearly evident in the shift of the GPC trace for the starting poly(ethylene glycol)-b-poly(styrene-cobenzocyclobutene), 15, (Mw ) 95 000, PDI ) 1.11) to that of the final hybrid nanoparticle-linear block copolymer, 16, (Mw ) 52 000, PDI ) 1.09) and demonstrates the controlled nature of this procedure. The solubility of these hybrid block copolymers were similar to that for the parent polystyrene nanoparticles which may be due to the relatively small size of the PEG block. Future work will examine the effect of larger PEG blocks on the solubility and phase behavior of hybrid PEG-PSt diblock copolymers. Conclusions

In conclusion, we have demonstrated a viable synthetic strategy for the controlled intramolecular collapse of linear (23) (a) Malenfant, P. R. L.; Groenendaal, L.; Fre´chet, J. M. J. J. Am. Chem Soc. 1998, 120, 10990. (b) Iyer, J.; Fleming, K.; Hammond, P. T. Macromolecules 1998, 31, 8758-8765. (c) Roma´n, C.; Fischer, H. R.; Meijer, E. W. Macromolecules 1999, 32, 5525-5531. (d) Gitsov, I.; Fre´chet, J. M. J. Macromolecules 1993, 26, 6536-6537.

8660 J. AM. CHEM. SOC.

9

VOL. 124, NO. 29, 2002

polymer chains to give single-molecule nanoparticles. By using a continuous addition strategy in conjunction with thermally activated benzocyclobutene-coupling chemistry, intermolecular cross-linking can be effectively eliminated even at high concentrations (0.1 M) which makes this a practical technique for the large-scale synthesis of well-defined nanoparticles. The size and cross-link density of the final nanoparticles can be accurately controlled by the initial degree of polymerization of the linear chain and the level of incorporation of the BCB coupling units. Conversion of the starting random coil linear polymer to the final nanoparticle results in a significant reduction in apparent molecular weight, up to 65-75%, which is fully consistent with the adoption of a dense, three-dimensional structure. The versatility of this approach is further demonstrated by the use of functionalized linear polymers and block copolymers to give reactive and architecturally complex nanoparticles, such as hybrid linear-nanoparticle copolymers. While the resulting nanoparticles share many features in common with dendritic macromolecules, it should be emphasized that a consequence of starting from preformed linear polymers is the ability to prepare larger molecular sizes more easily and the availability of the nanoparticles is greatly increased when compared to dendrimers. Acknowledgment. Financial support from the MRSEC Program of the National Science Foundation under Award Number DMR-9808677 for the Center for Polymeric Interfaces and Macromolecular Assemblies, NIST/ATP Cooperative Agreement No. 70NANB8H4013, and the IBM Corporation are gratefully acknowledged. We also thank Ron Huang, Stanford University for preliminary DLS measurements. JA026208X